Dating Apps Turn to Biometric Verification to Combat AI Deepfakes

In 2025, AI deepfakes flooded Tinder. According to the company, 98 percent of the content-moderation actions on the app were focused on combating fake accounts, scamming, and spam, pushing the company to expand its moderation efforts for its 60 million global users.

The problem is not unique to Tinder. The spread of AI-generated fake accounts has become a rising concern for dating apps, as things like chatfishing and voice cloning have gotten more widespread across the industry.

Among current online daters in the US, nearly half have been targeted by a dating scam, with US romance scams surpassing $4 billion between 2015 and 2025, according to the FBI.

The issue has forced just about every dating platform to adopt stricter authentication tools to ward off scammers. More and more, apps are asking that you surrender your most unique attribute in order to ensure that you’re a real person: your face.

More than a dozen major dating apps and sites use face recognition to verify users, often relying on liveness checks, 3D authentication, video selfies, or biometric ID to demonstrate humanness.

To combat the issue, Tinder launched the mandatory face-screening tool Face Check for new members in October 2025. “It’s not a feature. It’s a statement about what human connection online should require,” Yoel Roth, head of trust and safety for Match Group, which owns Tinder, said to WIRED in an email. Match Group tells WIRED that it is also starting to require Face Check for existing users in several of Tinder’s core markets, including the US and UK.

As face verification slowly transforms into a prerequisite for participation across the industry, the collective drive to combat bad actors has created a necessary and equally concerning new standard where “proof of human” profiles may soon be the norm.

Underpinning the shift, however, is a grim irony: In attempting to solve the problem of knowing whether someone is real, Big Dating is potentially creating more opportunities for privacy and security risks by forcing users to give up biometric data, while also legitimizing a culture of surveillance.

One company pushing for the use of ID-verification tech in dating apps is Tools for Humanity. The AI company created World ID, an iris-scanning biometric verification system, and partnered with Tinder in April, though it’s still optional for users. The most distinct feature about World ID is that confirmed users receive a “verified human” badge on the app after undergoing the process, which requires them to scan the colored part of their eye using the company’s orb device. They are given an identification number as their official World ID.

“In the near future, any dating app, like most online platforms in general, will need some sort of proof-of-human system like this,” says Tiago Sada, chief product officer at Tools for Humanity, adding that the way the majority of dating platforms handle verification is “super creepy, because you have to give your personal data to every single app that you want to use.” As stipulated by World’s policy, the orb does not share or retain any data; it is able to verify users without ever sending their data to a server. Once your image is confirmed, the orb sends the coded image to your phone before permanently deleting it, the company claims. Likewise, Face Check collects and stores an encrypted map of information about the shape of the user’s face, not the actual picture, Match said.

But even as dating apps claim that your biometric data is safe, security continues to be a significant concern.

Dating apps are subject to occasional data breaches, and while the inherent premise in these ID systems is that verification equals trust, cybersecurity researcher Katie Moussouris says that they are mostly a smoke screen. “Normalizing the idea that a face scan provides any assurance of identity or age is security theater.”

A liveness check requiring face scanning and live video only indicates a person was present at sign-up and does not guarantee that the person was not using AI to alter their appearance in real time. Likewise, an ID scan indicates an ID was presented, not that it belongs to the person who presented it. “Neither should be confused with trust,” Moussouris says. “A verified profile can still belong to a romance scammer who took over an account or created one using AI tools.”

And while passwords and usernames are interchangeable, faces are not, which makes them unique targets for attacks. More than that, biometric verification sets a risky precedent, says Woodrow Hartzog, a technology law professor at Boston University who previously served on the Massachusetts special commission on facial recognition.

“When we get accustomed to having our face scanned for everything we do online, it is really dangerous, not only potentially for the privacy risks but also to acclimate an entire population into thinking that that is normal behavior,” he says. “Facial recognition is the most dangerous surveillance tool ever created.”

Existing privacy laws are not equipped to deal with biometric verification on dating apps, according to Hartzog. He believes policies should ensure that less invasive options are used when they're available. In the case of biometrics, for example, your face, which can be monitored everywhere you go, is much more dangerous to collect than your thumbprint.

Utah’s State-Endorsed Digital Identity law, which went into effect in May, was one of the first major attempts at that. It limits the kinds of data that can be requested to only what is absolutely necessary while establishing a “duty of loyalty,” meaning platforms and providers cannot engage in practices that “conflict with the best interests of an individual” or “result in a disproportionate risk” to them. The law attempts to mitigate situations where people might have to present ID every time they log online, and also includes significant rules on the entity that’s collecting the data. The American Civil Liberties Union called it “the nation’s strongest digital identity bill.”

“Look at how easy it is to channel people's choices into agreeing to have their faces scanned as a condition of a good or service,” Hartzog says. “The more we can lean into policy frameworks that require strong protections from the company—rather than putting the onus on the person whose face is being scanned or who has to reveal their identity—then the better off I think we’ll be.”